6.05B: Enzyme Reaction Mechanisms - Quantiative Analyses of Serine Protease Catalysis

Search Fundamentals of Biochemistry

Learning Goals 

(Learning goals written by Claude, Sonnet 4.6, Anthropic)

Experimental Dissection of Serine Protease Mechanism

• Interpret kinetic data from substrate variation experiments to deduce the existence of a covalent acyl-enzyme intermediate in chymotrypsin catalysis — explaining why kcat for ester substrates is independent of leaving group quality (while uncatalyzed ester cleavage is not), why this implicates a rate-limiting deacylation step rather than leaving group departure, and how trapping of a ¹⁴C-labeled trimethylacetyl-enzyme adduct by gel filtration provides direct evidence for the covalent intermediate.
• Interpret pH-rate profiles to identify catalytic residues: a graph of kcat vs. pH showing half-maximum activity at pH ~6 implicates a general base His that must be deprotonated for activity, while a bell-shaped k_cat/K_M vs. pH curve additionally identifies a protonated N-terminal Ile (pKa ~10) required for a stabilizing salt bridge — and explain why kcat vs. pH informs on the chemical step while kcat/KM also reflects the substrate binding step.
• Describe the mechanistic evidence from chemical modification experiments — DIPF alkylating only Ser 195 among many serines (demonstrating hypernucleophilicity arising from the active site environment), tosyl-L-Phe-chloromethyl ketone modifying His in a 1:1 stoichiometry, and site-specific mutagenesis reducing activity to near-background when Ser 195 is replaced by Ala — and explain how these results, combined with conservation of Ser, His, and Asp in multiple protease sequences, converge to establish the catalytic triad.

Quantitative Deconstruction of Serine Protease Catalysis

• Describe how the ~10¹² overall rate enhancement by serine proteases relative to solution hydrolysis of peptide bonds (~17 kcal/mol) is partitioned among discrete physical and chemical factors — including general acid/base catalysis by His (~8.2 kcal/mol), electrostatic stabilization by the Asp of the triad (~0.8 kcal/mol), and a remaining ~8.1 kcal/mol attributed to ground state destabilization — and explain how the Brønsted linear free energy relationship (log k vs. log Ka) allows conversion of kinetic observations into ΔG contributions.
• Explain the concept of ground state destabilization as a catalytic mechanism — specifically how the Ser 195 nucleophilic oxygen in the ground state analog (GSA) complex is already 2.68 Å from the electrophilic carbonyl carbon (shorter than the van der Waals sum of ~3.2 Å), how the χ1 dihedral angle of Ser 195 is partially eclipsed (strained) in the GSA and rotates to relieve torsional strain in the transition state, and how the oxyanion hole N–H bonds to the substrate carbonyl O are nonplanar in the GSA (destabilizing) but optimally positioned in the transition state (stabilizing) — connecting these concepts to the principle that enzymes bind the transition state more tightly than the substrate.

Enzyme Catalysis in Organic Solvents

• Explain why chymotrypsin retains catalytic activity when suspended as a heterogeneous catalyst in anhydrous nonpolar organic solvents (e.g., octane) but not when dissolved in polar water-miscible solvents like DMSO — interpreting the kinetic (not thermodynamic) basis for stability in nonpolar solvents as restricted conformational mobility in the absence of water as plasticizer, and predicting that more hydrophobic solvents preserve a larger bound water monolayer and hence greater activity (k_cat/K_M is 15,000× higher in octane than in pyridine).
• Describe how enzymatic specificity is altered in nonpolar organic solvents — explaining why the k_cat/K_M advantage of Phe-esters over Ser-esters (50,000-fold in water) is reversed in octane (Ser-esters are 3-fold preferred) because the hydrophobic driving force for binding aromatic substrates is absent in a nonpolar solvent — and connect this to the practical applications of enzymes in organic synthesis, where their stereoselectivity, regioselectivity, and chemoselectivity can be exploited for transesterification, ammonolysis, and asymmetric synthesis not achievable in aqueous solution.

 

KristenProcko and HenryJakubowski have written this chapter section.

Introduction

To this point, we've presented mechanisms supported by PDB structures. However, much was known about enzyme mechanisms before ready access to the Protein Data Bank. Systematically, the kineticists, medicinal chemists, and molecular biologists (i.e., well-trained chemists) can change:

1. the substrate - for example, changing the leaving group or substituents of a hydrolyzable substrate;
2. the pH or ionic strength - which can give data about general acids/bases in the active site;
3. the enzyme - by chemical modification of specific amino acids or through site-specific mutagenesis;
4. the solvent - an odd idea on the surface, but it leads to new insights into enzyme catalysis.

We will focus on reaction mechanisms based on a mix of structural, kinetic, and thermodynamic data for the following enzymes, using these data to hypothesize a reaction mechanism consistent with the findings. Even with abundant data, there are often multiple proposed mechanisms for a given reaction. Kinetic data is vital as it can help to determine:

• the order of binding/dissociation of substrates and products;
• the rate constants for individual steps;
• and clues to the nature of catalytic groups found in the enzyme.

Chymotrypsin and Other Serine Proteases

Chymotrypsin (EC 3.4.21.1), an endoprotease, cleaves an internal peptide bond after aromatic side chains by hydrolysis. It also cleaves small ester and amide substrates after aromatic residues. For example, in Figure \(\PageIndex{1}\), cleavage occurs on the C-terminal side of the tyrosine residue, giving two peptide fragments.

Figure \(\PageIndex{1}\): Chymotrypsin cleavage of an example peptide substrate

Chymotrypsin has a similar mechanism to many other serine proteases that use the same catalytic triad, Ser 195, Asp 102, and His 95, so we'll study it in significant detail. In determining the mechanism of an enzyme, you have to change an experimental variable and see how catalytic activity changes. What can be changed? It turns out everything, including the solvent! Let's explore these changes and how they affect chymotrypsin activity.

1. Changing the substrate (for example, changing the leaving group or acyl substituents of a hydrolyzable substrate):

In the lab, studying the enzyme using small substrate mimics of a protein is easier than using a full protein substrate. The mimics include both esters and amides. Data from the small amide and ester substrates cleavage shown in Figure \(\PageIndex{2}\) suggest that a covalent intermediate is formed during chymotrypsin-catalyzed cleavage.

Figure \(\PageIndex{2}\): Small amide and ester substrates of chymotrypsin

Table \(\PageIndex{1}\) below shows kinetic data for the cleavage of these substrates.

Chymotrypsin substrate cleavage, 25 ^oC, pH 7.9
kinetic constants	Acetyl-Tyr-Gly-amide	Acetyl-Tyr-O Ethylester	Ester/Amide
kcat (s^-1)	0.50	193	390
Km (M)	0.023	0.0007	0.03
kcat/Km (M^-1s^-1)	22	280,000	12,700

Kinetic constants for chymotrypsin cleavage of N-acetyl-L-Trp Derivatives - N-acetyl-L-Trp-X
X	kcat (s^-1)	Km x 103 (M)
-OCH2CH3	27	0.097
-OCH3	28	0.095
-p-nitrophenol	31	0.002
-NH2	0.026	7.3

Table \(\PageIndex{1}\): Cleavage of peptides and ethylester substrate analogs by chymotrypsin

Here's how these data can be interpreted.

1. The k_cat and k_cat/K_m are larger and the K_m smaller for ester substrates compared to amide substrates, suggesting that amides are more challenging to hydrolyze (Table 2 above). This is expected given the amide's poorer leaving group.
2. The kcat for the hydrolysis of ester substrates doesn't depend on the nature of the leaving group (i.e., whether it is a poorer leaving group, such as methoxy, or a better leaving group, such as p-nitrophenolate), suggesting that this step isn't rate-limiting for ester cleavage. Without the enzyme, p-nitrophenyl esters are cleaved much more rapidly than methyl esters. Therefore, deacylation must be rate-limiting. But deacylation of what? If water were the nucleophile, releasing the leaving group would result in the simultaneous formation of both products, the free carboxyl group and the amine. Since they are not released simultaneously, this suggests the formation of an acyl-enzyme covalent intermediate.

A covalent intermediate can be trapped when the acyl end of the ester substrate is changed without changing the leaving group (a p-nitrophenyl group). Specifically, the deacylation of a trimethylacetyl group is much slower than that of an acetyl group. It is so slow that a ¹⁴C-labeled trimethylacetyl-labeled chymotrypsin intermediate can be isolated after incubation of chymotrypsin with ¹⁴C-labeled p-nitrophenyltrimethylacetate using gel filtration chromatography.

We have previously seen a kinetic mechanism consistent with these ideas. The data suggest a mechanism based on the chemical equations shown in Figure \(\PageIndex{3}\):

Figure \(\PageIndex{3}\): Chemical equations for chymotrypsin hydrolysis of a substrate involved a covalent intermediate with ping-pong kinetics.

In this reaction, a substrate S may bind to E to form a complex, which is then cleaved into products P and Q. Q is released from the enzyme, whereas P may remain covalently bound until it is expelled. This conforms exactly to the mechanism described above. For chymotrypsin-catalyzed cleavage, the step characterized by k₂ is the acylation step. The step characterized by k₃ is the deacylation step in which water attacks the acyl-enzyme to release product P (free phosphate in Lab 5). The mathematical equation for this reaction is shown below (without derivation)

\begin{equation}
\mathrm{v}_{0}=\frac{\left(\frac{\mathrm{k}_{2} \mathrm{k}_{3}}{\mathrm{k}_{2}+\mathrm{k}_{3}}\right) \mathrm{E}_{0} \mathrm{~S}}{\mathrm{~K}_{\mathrm{S}}\left(\frac{\mathrm{k}_{3}}{\mathrm{k}_{2}+\mathrm{k}_{3}}\right)+\mathrm{S}}
\end{equation}

For hydrolysis of ester substrates, which have better leaving groups than amides, deacylation is rate-limiting ( k₃<<k₂). As mentioned above, in amide hydrolysis, acylation can be rate-limiting (k₂ << k₃). From this, equation 6.5.1 can be simplified as shown in Table \(\PageIndex{2}\) below for ester and amide hydrolysis.

Ester hydrolysis (deacylation rate limiting, k3 << k2)	Amide hydrolysis (deacylation rate limiting, k2 << k3)
\begin{equation} \mathrm{v}_{0}=\frac{\mathrm{k}_{3} \mathrm{E}_{0} \mathrm{~S}}{\mathrm{~K}_{\mathrm{S}}\left(\frac{\mathrm{k}_{3}}{\mathrm{k}_{2}}\right)+\mathrm{S}} \end{equation}	\begin{equation} \mathrm{v}_{0}=\frac{\mathrm{k}_{2} \mathrm{E}_{0} \mathrm{~S}}{\mathrm{~K}_{\mathrm{S}}+\mathrm{S}} \end{equation}
\begin{equation} V_{M}=k_{3} E_{0} \end{equation}	\begin{equation} \mathrm{V}_{\mathrm{M}}=\mathrm{k}_{2} \mathrm{E}_{0} \end{equation}
\begin{equation} \mathrm{K}_{\mathrm{M}}=\mathrm{K}_{\mathrm{S}}\left(\frac{\mathrm{k}_{3}}{\mathrm{k}_{2}}\right) \end{equation}	\begin{equation} \mathrm{K}_{\mathrm{M}}=\mathrm{K}_{\mathrm{S}} \end{equation}

Table \(\PageIndex{2}\): Simplification of equation 6.5.1

As we saw before, for the rapid equilibrium assumption (when ES falls apart to E + S more quickly than it goes to the product, Chapter 6.3), K_M = Ks in the case of amide hydrolysis.

This reaction, with two reactants (bi) and two products (bi), involves a covalent enzyme intermediate in which the second reactant binds after the first product, Q, is released, and is called a BiBi Ping Pong reaction.  If k₂ >> k₃, an immediate and fast burst or release of product Q happens, followed by a slow release of P since the covalent E-P complex reacts with the second reactant with a low rate constant k₃.  When doing initial rate Michaelis-Menten kinetics, the initial velocity of Q formation, v₀, is not (dQ/dt)_t=0, but the slower constant rate after the burst phase, which is determined by k₃, the rate of cleavage of the E-P intermediate.     

The VCell computational model below shows the reaction BiBi Ping Pong reaction for the reaction involving an enzyme-P covalent intermediate.  The burst phases in Q are clearly seen if you rescale the graph as described below.

---

MODEL

BiBi-Ping Pong_Covalent Intermediate Irreversible reaction

Vcell reaction diagram (1-way arrows defined as reversible in actual mathematical model) and chemical equation

Yellow dots: Reaction Nodes (R1, R2 and R3 left to right)

Reaction made irreversible since k_r2 = 0, k_r3 = 0.

Initial parameter values:

• S₀ = 100, E₀ = 1, W (water in a hydrolysis reaction) = 50 and fixed throughout
• k_1f = 5, k_1r = 1, k_2f = 0.6,
• k_2f = 50, k_2r = 0
• k_3f = 0.05, k_3f = 0

Select Load [model name] below

 Select Start to begin the simulation.

1Interactive Element

Select Plot to change Y axis min/max, then Reset and Play  |  Select Slider to change which constants are displayed |  Select About  for software information.

To see the burst phase for reaction, change the time and parameters to these values:

• set Run time to 0.3
• Select Plot then Update Y axis max to 2
• Click Edit Plot Species and check just P and Q
• reset

Time course model made using Virtual Cell (Vcell), The Center for Cell Analysis & Modeling, at UConn Health.  Funded by NIH/NIGMS (R24 GM137787); Web simulation software (miniSidewinder) from Bartholomew Jardine and Herbert M. Sauro, University of Washington.  Funded by NIH/NIGMS (RO1-GM123032-04)

---

 

The burst phase is observed during ester hydrolysis, as described above, when k₂ >> k₃.

2. Changing the pH or ionic strength - which can give data about general acids/bases in the active site:

• a graph of k_cat as a function of pH indicates that a group of pK_a of approximately 6 must be deprotonated to express activity (i.e., V_max/2 is at about pH 6). This suggests that an active-site histidine is necessary, which, if it must be deprotonated to express activity, must act as a general base.
• a graph of k_cat/K_m shows a bell-shaped curve indicating the necessity of a deprotonated side chain with a pK_a of about 6 (i.e., the same His above) and a group that must be protonated with a pK_a of about 10. This is an N-terminal Ile in chymotrypsin, which must be protonated to form a stabilizing salt bridge in the protein. Note: This N-terminal Ile is actually at the 16 position in the inactive precursor of chymotrypsin (called chymotrypsinogen); upon activation of chymotrypsinogen, it loses the first 15 amino acids by selective proteolysis.

(Note: The PKAD is a database of experimentally measured pK_a values of protein ionizable groups. It is searchable by the PDB ID.)

3. Changing the enzyme - by chemical modification of specific amino acids, or through site-specific mutagenesis:

Here are some specific examples.

1. Modification of chymotrypsin (and many other proteases) with diisopropylphosphofluoridate (DIPF) modifies only one (Ser 195) of many serines in the protein, suggesting that it is hypernucleophilic and probably the amino acid that attacks the carbonyl C in the substrate, forming the acyl-intermediate. This reaction is illustrated in Figure \(\PageIndex{3}\). The figure also shows analogous molecules used in common insecticides, which act through a similar mechanism.

Figure \(\PageIndex{3}\): Mechanism of chymotrypsin inhibition by covalent modification by diisopropylphosphoflouridate.

2. Modification of the enzyme with tos-L-Phe-chloromethyl ketone inactivates the enzyme with a 1:1 stoichiometry, which results in a modified His, as shown in Figure \(\PageIndex{4}\).

Figure \(\PageIndex{4}\): Reaction of chymotrypsin and other serine proteases with chloromethyketones.

3. A comparison of the primary sequences of many proteases shows that three residues are invariant: a Ser, a His, and an Asp.
4. Site-specific mutagenesis shows that changing Ser 195 to Ala 195 reduces enzymatic activity to near background levels. This strongly suggests that Ser 195 is an active site nucleophile.

D. Changing the solvent. Yes, indeed, you can take chymotrypsin and show that it is active in anhydrous organic solvents. Surely this is impossible, you say! It is true, and we will explore it at the end of the chapter, since chymotrypsin activity in an aqueous solution is challenging enough to understand. No new chemistry is needed; it is just a change in what you can conceptualize.

Quantitative Analysis of Catalysis by Serine Proteases 

Based on Siyuan Du et al. Conformational ensembles reveal the origins of serine protease catalysis. Science 387, eado5068 (2025). DOI:10.1126/science.ado5068.

Our understanding of enzyme catalysis is incomplete. Other factors, in addition to those described in Chapter 6.1: How Enzymes Work, likely contribute to the 10¹² increase in the rate of serine protease-catalyzed hydrolysis of peptide bonds compared to solution hydrolysis. A quantitative analysis of each factor would help define its contribution and help find new ones. Siyuan et al. have done just that.

So far, we've seen that the active-site serine side chain (pKa ≈ 13, not considering its environment) acts as a nucleophile, assisted by a nearby general base/acid histidine (pKa ≈ 7), which, after proton abstraction to form His⁺, is stabilized by a nearby negatively charged aspartate. The developing δ^- on the carbonyl O in the transition state and the intermediate's full negative oxyanion are stabilized in the "oxyanion hole" through hydrogen bonds from N-H main chain donors, Ser 193 and Gly 195 (pKa ≈ 15). The reaction is intramolecular (entropically favored over bimolecular reactions) after substrate binding, except for the reaction with solvent water (55 M) needed to complete the reaction.  

Now compare this to the uncatalyzed hydrolysis in water.  H₂O (pKa ≈ 16) is the nucleophile/substrate, and other waters (pKa ≈ 16) stabilize the developing δ^-O in the transition state and the oxyanion intermediate. By analogy to the enzyme-catalyzed reaction, a stabilizing water molecule, which becomes, on accepting a proton, H₃O⁺ (pKa ≈ -2), increases the nucleophilicity of the "substrate" water by proton abstraction. 

Except for the adjacent His (pKa ≈ 7), which can more readily act as a general acid/base, there is nothing extraordinary about the groups involved in catalysis.  The constrained intramolecular reaction involving the positioning and likely distortion of substrates in the active site is a critical feature to quantify.  Another factor we haven't discussed is the reaction's dynamics.  The reaction has to proceed along reaction coordinates involving bond-making and breaking.  In addition, substrates, water, and side chains move into position for catalysis.  

Siyuan Du et al. show that fundamental concepts in chemistry and physics (torsion angles and strain, van der Waals interactions, hydrogen bonds, and entropy) can explain catalysis. They state that this "simplicity may inspire new ways of teaching enzyme catalysis, allowing instructors to reinforce the value of fundamental physical and chemical concepts and students to appreciate the tight connection between these concepts and the emergent, complex functions of biomolecules."  This is precisely what Fundamentals of Biochemistry attempts to do!  

They used molecular dynamics simulations and protein structures to quantitatively analyze catalytic contributions. Although we'll focus on serine proteases here, they extended their analyses to other enzymes.  They compared the PDB structures of 1231 wild-type serine proteases from 4 clans (structural superfamilies).  These structures included enzymes:

• without substrate or other ligands (i.e., the apo form of the enzyme);
• with bound ligand when the ligand structure is unperturbed (i.e., the ligand is a ground-state analog or GSA), which is typically a noncovalently-bound peptide or peptide analog);
• with covalently bound ligands resembling the transition state for a substrate (i.e., the ligand is a transition state analog or TSA).  The enzyme's active site would have a slightly altered structure to accommodate the tetrahedral TSA.  The TSAs are covalently attached to Ser 195 through an sp³ tetrahedral bond.  These TSAs include fluoromethylketones, peptide aldehydes, and boronic acids, which undergo nucleophilic addition at the analog's sp² carbonyl carbon or boron center as they form adducts.  The attached TSA stays covalently attached and does not react further.  The active site differs from the apo and GSA-bound form in the movement required to form the covalent adduct.  The active site is trapped in an "active" conformation.

These 1231 structures form a group, or "pseudo-ensemble," that collectively provides data that were hidden until the present analysis. These data describe subtle factors favoring catalysis and hint at the dynamical motion required within the active site, where groups move from the apo state to the transition/intermediate state. 

Quantum-mechanical analyses of reaction paths and dynamics for the hydrolysis of N-methylacetamide (NMA) in water (without enzyme) were conducted to provide a meaningful comparison with the solution-phase reaction.  (We saw NMA in Chapter 4.9: Protein Stability - Thermodynamics.) NMA mimics the peptide bond as shown below in Figure \(\PageIndex{5}\).  It likewise forms a tetrahedral intermediate during hydrolysis. 

Figure \(\PageIndex{5}\): Comparison of N-methylacetamide to a peptide bond.

Molecular dynamics simulations were used to obtain a distribution or ensemble of reacting molecules in solution to parallel the ensemble of protein structures.  

They then compared the distribution of molecules (solution) and side chains (enzyme) to compare the positioning of the serine side chain (compared to water), as catalysis has to involve its movement.  Three geometric variables were used: the distance (d_attack), angle (α_attack), and dihedral (Φ_attack). These parameters are described in Figure \(\PageIndex{6}\) below.

Figure \(\PageIndex{6}\): Visual representations of 3 key geometric parameters for hydrolysis of amide bonds.

Panel A shown d_attack, the distance between the nucleophile O on serine (or water in the uncatalyzed case) and the electrophilic C on the amide substrate, and α_attack, the angle between three atoms (the nucleophilic O, the electrophilic C, and the carbonyl O. The other key angle is a dihedral.  Panel B reviews examples of a dihedral angle.  The left Newman project shows the Φ dihedral, which is the angle of rotation around the C_b-C_c bond in the four atoms, three bonds C_a-C_b-C_c-C_d structure, where C_b is the front carbon, and C_c is the back carbon represented by the blue circle.  The right structure shows that the Φ_dihedral is also the angle of rotation around the red dotted line (---) connecting two gray planes.  Panel C shows the Φ_{attack.  In this} case, the gray plane contains the sp² hybridized amide substrate, and the red-dotted plane/triangle contains the plane defined by the nucleophilic O, the electrophilic C, and the carbonyl O.

The values for these parameters were very similar for all serine proteases studied, with mode (not median) values as follows: d_attack = 2.68 Å (SD = 0.14 Å), α_attack = 93° (SD = 7°), and Φ_attack = 84° (SD = 8°).  The distribution of water molecules for nucleophilic attack on DMA was much broader, with larger distances.  So, it seems that the serine in proteases has a shorter path and higher attack efficiency.  A caveat is that the active site can't be too rigid, which could hinder the motions required for catalysis.  Also, the serine must be pointed in the right direction.

To look for catalytic-specific motion, they compared torsion angles (backbone and side chain) for each residue in the enzyme with GSA and TSA bound. In addition, they measured changes between the apo and GSA-bound enzyme to assess changes in substrate binding alone.  Differences among these were interpreted as movement along the reaction pathway.  For trypsin, 32 torsion angle changes were observed upon substrate binding, and 23 upon nucleophilic attack (comparing TSA- and GSA-bound enzymes).  These changes were distributed throughout the enzyme.  Changes in the torsion angle χ¹ for the Cα–Cβ in the side chain of Ser 195 were large and occurred during substrate binding (−14°, leading to torsional strain on partial eclipsing) and the actual nucleophilic attack (+14°, relieving the torsional strain).  This large change was also seen for chymotrypsin and elastase (in the same clan).  Indeed, this common single change was the single largest torsional change for the three enzymes.

Figure \(\PageIndex{7}\) shows an interactive iCn3D model of the alignment of bovine pancreatic trypsin with a Kunitz Type serine protease Inhibitor-1 (3M7Q), a ground state noncovalent inhibitor, and trypsin with APA (1TPP), a covalent transition state analog (TSA). The gray backbone is the GS inhibitor complex, and the cyan backbone is the TSA analog covalent complex.  The inhibitors are not shown for clarity and simplicity.  Toggle the "A" key to switch between each form.

 Figure \(\PageIndex{7}\): Kunitz Type serine protease Inhibitor-1 (3M7Q), a ground state noncovalent inhibitor and trypsin with APA (1TPP), a covalent transition state analog. (Copyright; author via source).  Click the image for a popup or use this external link: https://www.ncbi.nlm.nih.gov/Structu...pUFy1femfcy4Y7

Look at the change in orientation of the serine O nucleophile.  Rotate the models to observe the distance and χ1 as the side-chain dihedral angle changes.

In the GSA-Trypsin, the nucleophilic Ser-O to the electrophilic carbonyl carbon in the planar sp² peptide bond is 2.7 Å (open this iCn3D link to see the distance).  The van der Waals radius for a carbonyl carbon (C) ≈ 1.7 Å, and ≈ 1.52 Å for an O, so the optimal distance is 3.2 Å.  Hence, the ground state structure is partially destabilized before bond-making actually occurs. The destabilization in the GSA contributes to catalysis.

A distance change occurs in the TSA complex as the serine oxygen moved ≈ 0.3 Å closer to the electrophilic carbon (now sp³), which is now raised 0.8 Å above the plane from its previous position in the sp² substrate since it is attached to the C. This move is expected during the reaction's dynamic progress as it proceeds along the reaction path. Now, the distance from the Ser-O to the electrophilic carbon in the covalent tetrahedral sp³ (nonplanar) TSA adduct is 1.6 Å (open this iCn3D link to see the distance).  (Note: the paper gives values of 1.6 in the text and 1.7 in a figure.) The net movement is ≈ 1.1 Å. 

How does this compare to the solution hydrolysis path?  Quantum-mechanical analyses show that the distance between the water O atom and the electrophilic C atom of the substrate in solution is 3.6 Å, exceeding the vdW distance.  This decreases  ≈ 2 Å, not ≈ 1.1 Å, as we saw for the enzyme reaction.  No bond rotations are noted.  Most of the changes arise from a simple translation of the nucleophilic O on the water, which is farther from the electrophilic carbon in the ground state.  Hence, the enzyme has a shorter reaction path, partly due to 1D bond dihedral rotation, than positioning the reactants via 3D translation.

Let's translate some of these changes into the thermodynamics parameter ΔG. Rates of catalysis can be related to the equilibrium constant, which is related to the ΔG⁰ for the reaction.  For example, suppose that breaking a strong bond is rate-limiting in a pathway, which often occurs in a thermodynamically stable molecule that reacts. In that case, the reaction rate also depends on the bond strength (i.e., a measure of its thermodynamic stability). For acids and bases, this is reflected in a linear relationship between log(k), the rate constant for the reaction, and the log(K_a), the equilibrium acid constant, as shown by the linear  Bronsted catalysis equation below:

\begin{equation}
\log k=\operatorname{alog}\left(K_{\mathrm{a}}\right)+C=-a p K_a+C
\end{equation}

ΔG for individual catalytic steps can be calculated using the example of a linear free-energy relationship.  For example, the observed 10⁶ increase in rate for trypsin, from histidine general base catalysis and its high effective intramolecular concentration, gives a value of ΔG around 8.2 kcal/mol, with an additional 0.8 kcal/mol from the stabilizing catalytic triad aspartate. The overall 10¹² increase in catalytic rate (net 17.1 kcal/mol) suggests another 8.1 kcal/mol remains unexplained.  The authors attribute the remaining catalysis to several factors.

One feature is based on the enzyme's ground state.  When the substrate binds, because the reaction proceeds intramolecularly, there is reduced conformational entropy and some unfavorable interactions, as described above.  These are decreased (relieved) in the transition site.  The net effect is another 7.6 kcal/mol of catalysis. They refer to these combined effects as ground-state destabilization.

Let's consider all of the factors they explored.  The first two deal with the entropy of positioning both the reactant and the active site serine:

• pre-orientation of the reactant in the active site: This makes it entropically easier to reach the "single" transition state than the reaction in water.
• pre-orientation of the enzyme serine nucleophile: Similarly, pre-orientation of the serine nucleophile in the vicinity of the substrate decreases the number of possible microstates needed to reach the transition state, entropically favoring the reaction.
• ground state destabilization of the complex through O-C bond shortening:  We mentioned above that the serine O to carbonyl C distance is shorter than the van der Waals interaction distance, which raises its energy as it moves into the direction of the transition state.
• ground state destabilization of the complex through partial eclipsing from the torsion angle of the catalytic serine:  As mentioned above, this rotation is reversed on conversion to the transition state. 
• ground state destabilization of the main oxyanion hole by hydrogen bonds:  Solution phase H bonds between amide Hs and carbonyl Os are more stable when planar. In the GSA, they are not planar and hence destabilized. 

The authors calculated  ΔG for each of these effects.  For an enzymatic reaction, the free energy of the transition state ΔG‡_solution > ΔG‡_enzyme.  The following equation describes ΔΔG‡:

\begin{equation}
\Delta \Delta \mathrm{G}^{\ddagger}=\Delta \mathrm{G}^{\ddagger}{ }_{\text {solution }}-\Delta \mathrm{G}^{\ddagger}{ }_{\text {enzyme }}
\end{equation}

Hence, the ΔΔG‡ value for the enzyme-catalyzed reaction and each contributing factor are all positive (>0). Figure \(\PageIndex{8}\) shows the individual contributions (combined to 16.6 kcal/mol) compared to the experimentally observed value of 17 kcal/mol.

Figure \(\PageIndex{8}\):  Data from Siyuan Du et al., ibid.  Error bars not included. 

Within error, catalysis can be described by destabilizing the ground state substrate and stabilizing the transition state based on principles you learned in introductory and organic chemistry: acid/base reactions, hydrogen bonds, entropy, torsional strain, bond angle strain, and van der Waals interactions.  The authors extended their analyses and found that the same factors account for catalysis by other enzymes of distinct folds using nucleophilic attack on carbonyl carbons.  These include lactamases, caspases, transcylases, acylases, peptidases, and other proteases.  They found similar structures and catalytic features (listed above) when aligning the PDB structures around the active-site nucleophile, oxyanion hole, oxyanion, and the substrates' electrophile carbon.  

Enzyme catalysis in organic solvents

In our earlier lists, we mentioned changing the solvent and exploring its effect on enzyme catalysis. It might seem a bit wild, but as we saw with the rhomboid protease, some enzymes work in hydrophobic environments. Also, lipases work at the boundary between the aqueous and hydrophobic worlds. For those interested, let's see what happens when we change solvents. These include putting the enzyme in various solvents or mixtures of solvents, as described below:

• Water-miscible solvents, such as ethanol and acetone, were added. If the water concentration was high enough, activity remained.
• Biphasic mixtures in which an aqueous solution of an enzyme was emulsified in a water-immiscible solvent like chloroform or ethyl acetate. The substrate would partition into both phases, while the product hopefully would end up in the organic phase.
• Nearly nonaqueous solvents, with a few % water at less than the solubility limits of water.
• Anhydrous organic solvents (0.01% water). This case is particularly astonishing, as enzymatic activity is often retained.

It is important to realize that in this last case, the enzyme is not in solution. Instead, it is in suspension and acts as a heterogeneous catalyst, much like palladium, which acts as a heterogeneous catalyst in the hydrogenation of alkenes. The suspension must be mixed vigorously and then sonicated to produce small suspended particles, so that diffusion of reactants into and out of the enzyme is not rate-limiting. Let's explore the activity of chymotrypsin in a nonpolar solvent.

Why aren't the enzymes inactive? Indeed, it must seem ridiculous that they aren't, since, as we learned earlier, proteins are marginally stable. On average, a 100-amino-acid protein is stabilized by only about 10 kcal/mol (41 kJ/mol) relative to the denatured state, or the equivalent of a few H bonds. Indeed, the hydrophobic effect, one of the dominant contributors to protein folding and stability, would not stabilize the native structure of enzymes in nonpolar organic solvents, and the protein would denature. It doesn't, however! Maybe the real question should be not whether water is necessary but how much water is essential. The enzyme can't "see" more than a monolayer or so of water around it. The data suggest that the nature of the organic solvent is very important. The most hydrophobic solvents are best for maintaining active enzymes! Chymotrypsin retains 10⁴ more activity in octane than in pyridine (see k_cat/K_m below), which is more hydrophilic than octane. The more polar the solvent, the more it can strip bound water away from the protein. If you add 1.5% water to acetone, the bound water increases from 1.2 to 2.4%, and the activity of chymotrypsin increases 1000-fold.

Table \(\PageIndex{3}\) below shows chymotrypsin activity in organic solvents.

Solvent	Structure	kcat/Km (M^-^1min^-1)	relative ratio kcat/Km	H2O bound to enzyme (%, w/w)
Octane		63	15000x	2.5
Toluene		4.4	1000x	2.3
Tetrahydrofuran		0.27	175x	1.6
Acetone		0.022	5.5x	1.2
Pyridine		<0.004	1x (.004)	1.0

Table \(\PageIndex{3}\): Chymotrypsin activity in organic solvents

Consider the following questions:

• How much water do the enzymes need? One chymotrypsin molecule in octane has fewer than 50 water molecules associated with it and retains some activity. About 500 water molecules are required to form a monolayer. Water can be added, presumably leading to more water binding and higher activity.
• How stable are the enzymes? Denaturation requires conformational flexibility, which requires water. The half-life of chymotrypsin in water at 60 ^oC is minutes, but in octane at 100 ^oC, it is hours. At 20 ^oC, the half-life in water is a few days, but in octane, it is greater than 6 months. Remember, two factors contribute to stability: 1. The protein can denature at high temperatures. 2. Chymotrypsin is a protease. It can cleave itself in an autoproteolytic reaction.

Table \(\PageIndex{4}\) below shows the half-life of chymotrypsin activity in water and octane

Solvent	60^oC	100^oC	20^oC
water	minutes	-	few days
octane	-	hours	> 6 months

Table \(\PageIndex{4}\): Half-life of chymotrypsin activity in water and octane at different temperatures

• Has the enzyme specificity changed? The net binding energy is a function of the substrate's binding energy minus the water's binding energy since water must be displaced from the active site on binding. In an anhydrous solvent, changes in specificity must be expected. For chymotrypsin, the driving force for binding substrates in water is primarily hydrophobic. In water, the k_cat/K_M for the reaction of N-acetyl-L-Ser esters is reduced by 50,000-fold compared to the Phe ester. However, chymotrypsin is three times more active toward Ser esters in octane than Phe esters.

Table \(\PageIndex{5}\) shows specificity changes in chymotrypsin in water and octane

Substrate	kcat/Km
	solvent: H2O	solvent: Octane
N-acetyl-L-Ser-ester	1x	3x
N-acetyl-L-Phe-ester	50,000x	1x

Table \(\PageIndex{5}\): Specificity changes in chymotrypsin in water and octane

Now, consider competitive inhibitors. Naphthalene binds 18 times more tightly than 1-naphthoic acid, but chymotrypsin binds naphthoic acid 310 times as tightly in octane. Likewise, the ratio of [k_cat/K_m (L isomer)]/[k_cat/K_m (D isomer)] of N-acetyl-D- or N-acetyl-L-Ala-chloroethyl esters is 1000-10,000 in water, but less than 10 in octane.

Table \(\PageIndex{6}\) shows chymotrypsin inhibition constants in water and octane.

Inhibitor	Inhibition Constant Ki (nM)
	In water	In Octane
Benzene	21	1000
Benzoic acid	140	40
Toluene	12	1200
Phenylacetic acid	160	25
Naphthalene	0.4	1100
1-Naphthoic acid	7.2	3

Table \(\PageIndex{6}\): Chymotrypsin inhibition constants in water and octane

Can new reactions be carried out in nonpolar solvents? The quick answer is yes since reactions in aqueous solutions can be unfavorable due to low K_eq values, side reactions, or insolubility of reactants. Consider lipases, which hydrolyze fatty acid esters in aqueous solutions. In nonaqueous solutions, reactions such as transesterification or ammonolysis can be performed.

Enzymes are active in organic solvents, which contradicts our central concepts of protein stability. Two reasons could explain this stability:

1. From a thermodynamic perspective, the enzyme may be stable in organic solvents. However, as discussed above, this is inconceivable given the delicate balance of noncovalent and hydrophobic interactions required for protein stability.
2. The second reason must prevail: the protein cannot unfold from a kinetic standpoint. Conformational flexibility is required for denaturation, which requires water as the solvent. Denaturation in organic solvents is kinetically, not thermodynamically, controlled.

A specific example helps illustrate the effects of different solvents on chymotrypsin activity. Dry chymotrypsin can be dissolved in DMSO, a water-miscible solvent. In this solvent, chymotrypsin is entirely and irreversibly denatured. No activity is observed when next diluted 50X with acetone containing 3% water. (In the final dilution, the concentrations of solvents are 98% acetone, 2.9% water, and 2% DMSO.) However, the enzyme is very active if dry chymotrypsin is added to a mixture of 98% acetone, 2.9% water, and 2% DMSO. We end up with the same final solvent state, but the enzyme has no activity in the first case, while in the second case, it retains activity. These ideas are illustrated in Figure \(\PageIndex{9}\).

Figure \(\PageIndex{9}\): Chymotrypsin activity in acetone depends on the order of solvent addition

Dry enzymes added to a concentrated water-miscible organic solvent (like DMSO) will dissolve and surely denature, but will retain activity when added to a concentrated water-immiscible solvent (like octane), in which the enzyme will not dissolve but stay in suspension.

It appears the enzymes have very restricted conformational mobility in nonpolar solvents. By lyophilizing (freeze-drying) the enzyme against a specific ligand, a given conformation of a protein can be trapped or imprinted onto the enzyme. For example, if the enzyme is dialyzed against a competitive inhibitor (which can be extracted by the organic solvent), freeze-dried to remove water, and then added to a nonpolar solvent, the enzyme activity of the "imprinted" enzyme in nonpolar solvents is as much as 100x as great as when no inhibitor was present during the dialysis. Suppose chymotrypsin is lyophilized from solutions of different pHs. In that case, the resulting V_M/K_M curve for ester hydrolysis in octane is bell-shaped, with the initial rise in activity reaching half-maximum at a pH of around 6.0 and the subsequent fall in activity reaching half-maximum at a pH of approximately 9.

The use of enzymes in organic solvents allows new routes to organic synthesis. Enzymes, which are so helpful in synthetic reactions, are:

• stereoselective - can differentiate between enantiomers and between prochiral substrates
• regioselective - can differentiate between identical functional groups in a single substrate
• chemoselective - can differentiate between different functional groups in a substrate (such as between a hydroxyl group and an amine for an acylation reaction)

Enzymes in anhydrous organic solvents are useful (from a synthetic point of view) not only because they can catalyze new types of reactions (such as transesterification, ammonolysis, and thiolysis) but also because the stereoselectivity, regioselectivity, and chemoselectivity of the enzyme often differ from those of the enzyme in water.

Organic reactions are usually conducted in organic solvents because many organic molecules react with water, and the reagents and products are generally insoluble in water. In a manner analogous to using an enzyme as a heterogeneous catalyst in a nonpolar solvent, Sharpless is pioneering a technique to conduct organic reactions in water. They (Narayan et al.) have shown that many unimolecular and bimolecular reactions occur faster in water than in organic solvents. As in enzyme catalysis in nonpolar solvents, reactions in water must be mixed vigorously to disperse reactants into microdroplets (a suspension), significantly increasing the surface area available for water to act on transition states or intermediates and stabilize them through hydrogen bonding. They called these "on water" reactions because the reactants usually float on the water. Using this process, they have performed cycloadditions, alkene reactions, Claisen rearrangements, and nucleophilic substitution reactions. One cycloaddition reaction went to completion in ten minutes at room temperature, compared to 18 hours in methanol and 120 in toluene. Adding nonpolar solvents at certain times significantly increased the reaction rate.

Summary

(Summary written by Claude, Sonnet 4.6, Anthropic)

This chapter integrates kinetic, chemical modification, structural, and computational approaches to establish the mechanism of chymotrypsin in quantitative detail — and then extends the analysis to the counterintuitive phenomenon of enzyme catalysis in nonpolar organic solvents.

Systematic experimental dissection of chymotrypsin's mechanism illustrates the general strategy for establishing enzyme mechanisms before crystal structures are available. Substrate variation experiments reveal the existence and chemical identity of a covalent acyl-enzyme intermediate. The key evidence: kcat for ester hydrolysis is identical regardless of leaving group quality (methoxy vs. p-nitrophenolate), while the uncatalyzed hydrolysis of p-nitrophenyl esters is much faster than methyl esters. This dissociation of leaving group effect from kcat indicates that leaving group departure is not rate-limiting for esters — deacylation (water attacking the acyl-enzyme) is the slow step. The release of the two products (amine and carboxylate) is nonsimultaneous, further supporting a covalent intermediate. Direct trapping confirms this: the bulky trimethylacetyl group deacylates so slowly that ¹⁴C-labeled trimethylacetyl-chymotrypsin can be isolated intact by gel filtration, demonstrating the intermediate's existence. For amide substrates (poorer leaving groups), the acylation step (k₂) is rate-limiting because the amine is a worse leaving group than an alcohol, making k₂ << k₃; for esters (better leaving groups), deacylation is rate-limiting because k₃ << k₂. This reaction — two substrates (S and H₂O), two products (Q and P), with a covalent enzyme intermediate, in which the first product Q leaves before the second substrate H₂O binds — is a BiBi Ping Pong mechanism. When k₂ >> k₃, an initial burst phase of rapid Q formation is observed as the acyl-enzyme accumulates, followed by the slower steady-state phase governed by k₃; true initial rate kinetics should be taken from the steady-state phase, not the burst.

pH-rate profiles reveal the ionization states of catalytic residues. The kcat vs. pH curve shows half-maximal activity at pH ~6, indicating that a residue with pKa ~6 (His 57) must be deprotonated (acting as a general base) for catalysis. The bell-shaped kcat/KM vs. pH curve shows the same His requirement and an additional requirement for a protonated group at pKa ~10 — the N-terminal Ile 16 (generated by activation of chymotrypsinogen), which must be protonated to form a stabilizing salt bridge within the protein. Chemical modification experiments converge on the same active site residues: DIPF phosphorylates only Ser 195 among the many serines in chymotrypsin, demonstrating that the active site environment makes this serine hypernucleophilic relative to surface serines; tosyl-L-Phe-chloromethyl ketone alkylates His 57 with 1:1 stoichiometry; and Ser 195 → Ala mutagenesis abolishes activity. Together with sequence conservation of Ser, His, and Asp across diverse serine proteases, these results firmly establish the Ser 195–His 57–Asp 102 catalytic triad as the mechanistic core.

Quantitative deconstruction of the ~10¹² rate enhancement (equivalent to ~17 kcal/mol of ΔΔG‡) by serine proteases over solution amide hydrolysis, developed by Siyuan Du et al. through analysis of 1231 PDB structures across four serine protease clans, partitions the catalytic power among discrete contributions that can each be estimated using the Brønsted linear free energy relationship. General acid/base catalysis by His (~8.2 kcal/mol) and Asp stabilization of protonated His (~0.8 kcal/mol) account for only ~9 kcal/mol — leaving ~8 kcal/mol unexplained by classical mechanisms. The remaining contribution arises from ground state destabilization: in ground state analog (GSA) complexes, the Ser 195 Oγ is already closer to the electrophilic carbonyl carbon than van der Waals contact distance (~2.68 Å vs. ~3.2 Å sum of vdW radii), partially destabilizing the ground state; the χ₁ dihedral angle of Ser 195 is in a partially eclipsed (torsionally strained) conformation in the GSA that rotates by ~14° to relieve strain in the transition state; and the N–H donors of the oxyanion hole (backbone NHs of Gly 193 and Ser 195) are nonplanar with the substrate carbonyl O in the GSA (destabilizing the H-bond) but achieve optimal geometry in the transition state. Together, these contribute ~7.6 kcal/mol. The combination of pre-orientation (reducing the translational and rotational entropy cost of reaching the transition state from a broad ensemble of solution configurations to the narrow ensemble in the active site), ground state destabilization, and transition state stabilization through acid/base catalysis, the oxyanion hole, and the catalytic triad collectively account for virtually the entire observed rate enhancement. Importantly, the enzyme achieves this partly through a shorter reaction path (~1.1 Å translation of Ser O to C, compared to ~2 Å for water O to C in solution), enabled by the 1D torsional rotation of Ser 195 rather than 3D diffusional repositioning.

Enzyme catalysis in organic solvents provides a striking demonstration that water is not mechanistically essential for chymotrypsin activity, though it is kinetically required for conformational flexibility. When added as a suspension to anhydrous hydrophobic solvents like octane, chymotrypsin retains significant activity — but in polar water-miscible solvents like DMSO, it denatures irreversibly. The order of addition is critical: dry enzyme added directly to 98% acetone/3% water/2% DMSO is active, but enzyme first dissolved in DMSO and then diluted with acetone/water is inactive. These observations establish that denaturation in organic solvents is kinetically controlled (conformational flexibility required for unfolding is absent when water is absent) rather than thermodynamically controlled. Enzyme stability is dramatically enhanced: the half-life in octane at 100°C is hours, compared to minutes in water at 60°C. Importantly, substrate specificity is fundamentally altered in organic solvents: in octane, where there is no hydrophobic driving force for aromatic substrate binding, Ser-esters are preferred over Phe-esters (3-fold), reversing the 50,000-fold selectivity for Phe-esters seen in water. This substrate memory and altered specificity in nonpolar solvents enables practical applications including stereoselective, regioselective, and chemoselective transesterification, ammonolysis, and thiolysis reactions that are not feasible in aqueous solution — making enzyme catalysis in organic solvents an important tool in pharmaceutical and industrial organic synthesis.